The diprate R package

You can install the diprate R package by following the instructions here: https://github.com/QuLab-VU/dipDRC

Combining VLW001 and VLW_V23b datasets

After having fixed the apparent error due to incorrect assignment of missed/temporary barcodes, the VLW001 dataset looks appropriately clean for combining with the VLW_V23b dataset to enable analysis of all four cell line simultaneously and consider analyzing the 3D data.

require(diprate)
Loading required package: diprate
source("../functions/namingStandards.r")
keep_cols <- c(std_colnames,c("ch2.pos","orig.cell.count","file_name"))
d1 <- read.csv("../data/VLW001_dataset.csv", as.is=TRUE)
d2 <- read.csv("../data/VLW_V23b_dataset.csv", as.is=TRUE)
d <- merge(d1,d2, all=TRUE)

Standardize data

The VLW001 dataset did not have a calculation for viable cells (cell.count - ch2.pos) and several data types with NA values.

Steps to perform:

  • Remove extra columns
  • Calculate viable cells (cell.count <- orig.cell.count - ch2.pos)
  • Replace NA in drug1 as control and drug1.conc as 0
  • Calculate fraction of dead cells (dead.cell.frxn <- ch2.pos/orig.cell.count)
  • Identify high-density culture (confluence) and remove time points once reached
d[is.na(d$orig.cell.count),'orig.cell.count'] <- d[is.na(d$orig.cell.count),'cell.count']
cc <- d[d$orig.cell.count==d$cell.count,'cell.count'] - 
    d[d$orig.cell.count==d$cell.count,'ch2.pos']
d[d$orig.cell.count==d$cell.count & !is.na(d$cell.count),'cell.count'] <- cc[!is.na(cc)]
rm(cc)
d <- d[,keep_cols]
d[is.na(d$drug1.conc),'drug1.conc'] <- 0
d[is.na(d$drug1),'drug1'] <- 'control'
d <- d[!is.na(d$cell.count),]
d$drug1.units <- 'M'
d <- d[order(d$plate.name,d$uid,d$time),]
d$cell.count[d$cell.count<=0] <- 1

rownames(d) <- NULL

Modified from VLW_V23b_analysis.Rmd
Should exclude wells O14–O20 on drug plate 3 since it appears they really contain drugs.

d <- d[!(d$well %in% paste0('O',14:20) & grepl("-D3", d$plate.name)),]

Subset 2D cultures only.

a <- d[grepl("^V2", d$plate.name),]

Perform QC on controls.

par(mfrow=c(2,2))
ctrl <- sapply(unique(a$cell.line), function(cl) {
    controlQC(  a[a$cell.line==cl & a$drug1.conc==0,]$time,
                a[a$cell.line==cl & a$drug1.conc==0,]$cell.count,
                a[a$cell.line==cl & a$drug1.conc==0,]$uid,
                cell.line.name=cl,
                min.ar2=0.95,
                plotIt=TRUE,
                ret.type="all")
    }, simplify=FALSE)

Determine maximum number of cells consistent with exponential growth.

max.counts <- sapply(unique(a$cell.line), function(cl) {
    findMaxCounts(  a[a$cell.line==cl & a$drug1.conc==0,]$time,
                a[a$cell.line==cl & a$drug1.conc==0,]$cell.count,
                a[a$cell.line==cl & a$drug1.conc==0,]$uid,
                min.ar2=0.95)}, 
    simplify=FALSE)
The first 5 pts of ids = V2-11T-D1_B02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-11T-D1_I02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-11T-D1_O21 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-11T-D2_K13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_B02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_C02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_C13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_D13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_E02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_E13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_F13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_G13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_H02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_H13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_I14 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_I15 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_I16 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_I21 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_J13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_K13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_N02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_N13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D2_O13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_B02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_B13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_C13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_D02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_D13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_E02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_E13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_F13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_G02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_G13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_H13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_I02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_I13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_J02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_J13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_K02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_K13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_L02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_L13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_M13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_N02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_N13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_O02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_O13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_O21 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_O22 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D3_O23 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_B02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_B13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_C02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_C13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_D02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_D13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_E02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_E13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_F02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_F13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_G02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_G13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_H02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_H13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I14 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I16 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I17 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I18 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I19 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I20 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I21 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I22 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_I23 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_J02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_J13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_K02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_K13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_L02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_L13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_M02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_M13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_N02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_N13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_O02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D4_O13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_B02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_B13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_C02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_C13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_D02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_D13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_E02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_E13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_F02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_F13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_G02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_G13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_H02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_H13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_I02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_I13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_J02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_J13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_K02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_K13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L14 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L15 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L16 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L17 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L18 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L19 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L20 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L21 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L22 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_L23 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_M02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_M13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_N02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_N13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_O02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-16T-D5_O13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_B02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_D13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_F02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_G02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_H02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_I02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_J02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_J13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_K02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_K13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_L02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_L13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_M13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_N02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_N13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_O15 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_O16 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_O18 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_O20 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D1_O22 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D2_I18 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D2_I22 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D3_G13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D3_H13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D3_I13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D3_O22 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D4_C13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D4_D02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-21T-D4_I22 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_B13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_C02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_C13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_D02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_E02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_E13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_G02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_H02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_I02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_I13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_J02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_K02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_K13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_L02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_L13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_M02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_M13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_N02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_N13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_O02 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_O13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_O14 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_O19 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_O20 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D1_O21 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D2_B13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D2_D13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D2_I16 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D2_I22 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D2_K13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D2_N13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D3_H13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D4_L13 is not linear within ar2 = 0.95
The first 5 pts of ids = V2-29T-D4_M02 is not linear within ar2 = 0.95

Show histogram of max counts for each cell line.

par(mfrow=c(1,4))
invisible(lapply(names(max.counts), function(n) boxplot(max.counts[[n]]$max.exp.counts, ylim=c(0,2000), main=n)))

QC of 16T data

Appears that cells are reaching confluence during the experiment only for 16T. Identify the max counts where growth has not yet been restricted and use to obtain a reasonable max value to limit data.

Remove data from 16T when cell count exceeds the max consistent with exponential growth in the control wells.

max.16T <- quantile(max.counts[['16T']]$max.exp.counts, 0.05)
uid2filter <- unique(d[d$cell.count>max.16T,'uid'])
times2filter <- sapply(uid2filter, function(id) min(d[d$uid==id & d$cell.count>max.16T,'time']))

b <- d[d$uid %in% uid2filter,]
b <- do.call(rbind, lapply(seq_along(uid2filter), function(i) 
    b[b$uid==uid2filter[i] & b$time <times2filter[i],]))

d <- d[!d$uid %in% uid2filter,]
d <- rbind(d,b)
d <- d[order(d$plate.name,d$uid,d$time),]
a <- d[grepl("^V2", d$plate.name),]
par(mfcol=c(5,4))
invisible(lapply(unique(a$plate.name), function(mypn) do.call(plotGC,append(
    list(ylim=c(0,5), main=mypn, leg=FALSE),getGCargs(a[a$plate.name==mypn & a$drug1.conc==0,])))))

Remove bad controls

Use a 95% confidence interval to determine which wells may be outliers. Use code in controlQC function in diprate package.

ctrl_ids <- unique(a[a$drug1.conc==0,'uid'])

ctrlQC <- function(dat)
{
    ids <- as.character(unique(dat$uid))
    
    m <- lm(log2(cell.count) ~ time * uid, data=dat)
    rates <- coef(m)[grepl("time", names(coef(m)))]
    rates <- c(rates[1], rates[-1] + rates[1])
    ids.ok <- names(rates[rates < mean(rates) + sd(rates) & rates > 
        mean(rates) - sd(rates)])
    ids.ok <- gsub("time:uid", "", ids.ok)
    ids.ok <- sub("time", ids[1], ids.ok)
    out <- list(good_controls=ids.ok, outliers=setdiff(ids,ids.ok))
    return(out)
}

Assess controls and remove outliers from data.

ctrl_quality <- lapply(unique(a$cell.line), function(cl) ctrlQC(a[a$cell.line==cl & a$drug1.conc==0,]))
bad_ctrls <- unlist(sapply(ctrl_quality, '[[', 'outliers'))
d <- d[!d$uid %in% bad_ctrls,]
a <- d[grepl("^V2", d$plate.name),]
par(mfcol=c(5,4))
invisible(lapply(unique(a$plate.name), function(mypn) do.call(plotGC,append(
    list(ylim=c(0,5), main=mypn, leg=FALSE),getGCargs(a[a$plate.name==mypn & a$drug1.conc==0,])))))

Examine growth curves

Use a single drug to assess growth curves across all drug concentrations for each cell line.

par(mfrow=c(1,4))
invisible(lapply(unique(a[a$drug1=="VU0823511", 'plate.name']), function(pn) do.call(plotGC,append(
    list(ylim=c(-2,5), main=pn, leg=FALSE),getGCargs(a[a$plate.name==pn & a$drug1=="VU0823511",])))))

Replace drug names

Obtained, from Dave Westover, drug names to replace VU-specific identifiers.

new_drug_names <- readxl::read_xlsx("../data/VLW001_full_drugnames.xlsx")

Plan is to replace drug names by matching d$drug1 to new_drug_names$VU_ID and replace with new_drug_names$"Chemical Name". However, the chemical names need to be cleaned.

new_drugs <- as.character(new_drug_names$"Chemical Name")

Algorithm:

  • split on (, keeping first part.
  • remove dashes
  • remove spaces if before numbers
  • split on spaces, keeping first part (remove salts)
  • If only first letter cap or if != “Ro”, make lowercase
temp <- sapply(strsplit(new_drugs, " (", fixed=TRUE), "[[", 1, simplify=TRUE)
temp[grep(" [0-9]",temp)] <- gsub(" ", "", temp[grep(" [0-9]",temp)])
temp <- gsub("-", "", temp)
temp <- sapply(strsplit(temp, " "), "[[", 1, simplify=TRUE)
temp[!grepl("Ro", temp) & grepl("^[A-Z][a-z]", temp)] <- 
    tolower(temp[!grepl("Ro", temp) & grepl("^[A-Z][a-z]", temp)])
new_drug_names$drug1 <- temp

Replace d$drug1 values starting with VU with appropriate value in new_drug_names$drug1 then subset data for 2D data only and save into object a.

Exclude file_name from Thunor data.

d[grep("^VU",d$drug1), 'drug1'] <- 
    new_drug_names[match(d[grep("^VU",d$drug1),'drug1'], new_drug_names$VU_ID),]$drug1

a <- d[grepl("^V2", d$plate.name),]
a <- a[,colnames(a)!="file_name"]

Save 2D data for Thunor

OVERWRITE <- FALSE
thunor_2D_dat <- a
thunor_2D_dat$upid <- paste(thunor_2D_dat$expt.date, thunor_2D_dat$plate.name, sep='_')
thunor_2D_dat$drug1.units <- 'M'
thunor_fn <- "../data/VLW_V23ab_2Donly_Thunor_dataset.csv"
if(!file.exists(thunor_fn) | OVERWRITE) write.csv(thunor_2D_dat, file=thunor_fn, row.names=FALSE)

Save all data (including file name with full path)

alldat_fn <- "../data/VLW_V23ab_dataset.csv"
if(!file.exists(alldat_fn) | OVERWRITE) write.csv(d, file=alldat_fn, row.names=FALSE)
---
title: "Analysis of VLW 2Dvs3D combined data"
output: 
    html_notebook:
        self_contained: yes
author: "Darren R Tyson"
date: "2020-11-17"
---

## The `diprate` R package
You can install the `diprate` R package by following the instructions here: https://github.com/QuLab-VU/dipDRC  


### Combining VLW001 and VLW_V23b datasets
After having fixed the apparent error due to incorrect assignment of missed/temporary barcodes, the VLW001 dataset looks appropriately clean for combining with the VLW_V23b dataset to enable analysis of all four cell line simultaneously and consider analyzing the 3D data.

```{r Setup}
require(diprate)
source("../functions/namingStandards.r")
keep_cols <- c(std_colnames,c("ch2.pos","orig.cell.count","file_name"))
```

```{r Load data}
d1 <- read.csv("../data/VLW001_dataset.csv", as.is=TRUE)
d2 <- read.csv("../data/VLW_V23b_dataset.csv", as.is=TRUE)
d <- merge(d1,d2, all=TRUE)
```


#### Standardize data
The VLW001 dataset did not have a calculation for viable cells (`cell.count - ch2.pos`) and several data types with NA values.

Steps to perform:  

* Remove extra columns
* Calculate viable cells (`cell.count <- orig.cell.count - ch2.pos`)
* Replace NA in `drug1` as `control` and `drug1.conc` as `0`
* Calculate fraction of dead cells (`dead.cell.frxn <- ch2.pos/orig.cell.count`)
* Identify high-density culture (confluence) and remove time points once reached

```{r Quantify viable cells}
d[is.na(d$orig.cell.count),'orig.cell.count'] <- d[is.na(d$orig.cell.count),'cell.count']
cc <- d[d$orig.cell.count==d$cell.count,'cell.count'] - 
    d[d$orig.cell.count==d$cell.count,'ch2.pos']
d[d$orig.cell.count==d$cell.count & !is.na(d$cell.count),'cell.count'] <- cc[!is.na(cc)]
rm(cc)
```


```{r Remove extra columns}
d <- d[,keep_cols]
d[is.na(d$drug1.conc),'drug1.conc'] <- 0
d[is.na(d$drug1),'drug1'] <- 'control'
d <- d[!is.na(d$cell.count),]
d$drug1.units <- 'M'
d <- d[order(d$plate.name,d$uid,d$time),]
d$cell.count[d$cell.count<=0] <- 1

rownames(d) <- NULL
```

*Modified from `VLW_V23b_analysis.Rmd`*  
Should exclude wells O14–O20 on drug plate 3 since it appears they really contain drugs.
```{r}
d <- d[!(d$well %in% paste0('O',14:20) & grepl("-D3", d$plate.name)),]
```

Subset 2D cultures only.
```{r Subset 2D}
a <- d[grepl("^V2", d$plate.name),]
```

Perform QC on controls.
```{r Control QC, fig.height=8, fig.width=8, message=FALSE, warning=FALSE}
par(mfrow=c(2,2))
ctrl <- sapply(unique(a$cell.line), function(cl) {
	controlQC(	a[a$cell.line==cl & a$drug1.conc==0,]$time,
				a[a$cell.line==cl & a$drug1.conc==0,]$cell.count,
				a[a$cell.line==cl & a$drug1.conc==0,]$uid,
				cell.line.name=cl,
			    min.ar2=0.95,
				plotIt=TRUE,
                ret.type="all")
	}, simplify=FALSE)
```

Determine maximum number of cells consistent with exponential growth.
```{r Get max counts, message=FALSE, warning=FALSE}
max.counts <- sapply(unique(a$cell.line), function(cl) {
	findMaxCounts(	a[a$cell.line==cl & a$drug1.conc==0,]$time,
				a[a$cell.line==cl & a$drug1.conc==0,]$cell.count,
				a[a$cell.line==cl & a$drug1.conc==0,]$uid,
                min.ar2=0.95)}, 
	simplify=FALSE)

```

Show histogram of max counts for each cell line.
```{r fig.height=2, fig.width=8}
par(mfrow=c(1,4))
invisible(lapply(names(max.counts), function(n) boxplot(max.counts[[n]]$max.exp.counts, ylim=c(0,2000), main=n)))
```

#### QC of 16T data
Appears that cells are reaching confluence during the experiment only for 16T. Identify the max counts where growth has not yet been restricted and use to obtain a reasonable max value to limit data.  

Remove data from 16T when cell count exceeds the max consistent with exponential growth in the control wells. 
```{r Remove data after max counts}
max.16T <- quantile(max.counts[['16T']]$max.exp.counts, 0.05)
uid2filter <- unique(d[d$cell.count>max.16T,'uid'])
times2filter <- sapply(uid2filter, function(id) min(d[d$uid==id & d$cell.count>max.16T,'time']))

b <- d[d$uid %in% uid2filter,]
b <- do.call(rbind, lapply(seq_along(uid2filter), function(i) 
    b[b$uid==uid2filter[i] & b$time <times2filter[i],]))

d <- d[!d$uid %in% uid2filter,]
d <- rbind(d,b)
d <- d[order(d$plate.name,d$uid,d$time),]
a <- d[grepl("^V2", d$plate.name),]
```

```{r fig.height=11, fig.width=8}
par(mfcol=c(5,4))
invisible(lapply(unique(a$plate.name), function(mypn) do.call(plotGC,append(
    list(ylim=c(0,5), main=mypn, leg=FALSE),getGCargs(a[a$plate.name==mypn & a$drug1.conc==0,])))))
```

#### Remove bad controls
Use a 95% confidence interval to determine which wells may be outliers. Use code in `controlQC` function in `diprate` package.
```{r}
ctrl_ids <- unique(a[a$drug1.conc==0,'uid'])

ctrlQC <- function(dat)
{
    ids <- as.character(unique(dat$uid))
    
    m <- lm(log2(cell.count) ~ time * uid, data=dat)
    rates <- coef(m)[grepl("time", names(coef(m)))]
    rates <- c(rates[1], rates[-1] + rates[1])
    ids.ok <- names(rates[rates < mean(rates) + sd(rates) & rates > 
        mean(rates) - sd(rates)])
    ids.ok <- gsub("time:uid", "", ids.ok)
    ids.ok <- sub("time", ids[1], ids.ok)
    out <- list(good_controls=ids.ok, outliers=setdiff(ids,ids.ok))
    return(out)
}


```


Assess controls and remove outliers from data.
```{r}
ctrl_quality <- lapply(unique(a$cell.line), function(cl) ctrlQC(a[a$cell.line==cl & a$drug1.conc==0,]))
bad_ctrls <- unlist(sapply(ctrl_quality, '[[', 'outliers'))
d <- d[!d$uid %in% bad_ctrls,]
a <- d[grepl("^V2", d$plate.name),]
```

```{r fig.height=11, fig.width=8}
par(mfcol=c(5,4))
invisible(lapply(unique(a$plate.name), function(mypn) do.call(plotGC,append(
    list(ylim=c(0,5), main=mypn, leg=FALSE),getGCargs(a[a$plate.name==mypn & a$drug1.conc==0,])))))
```

#### Examine growth curves
Use a single drug to assess growth curves across all drug concentrations for each cell line.
```{r Check drug QC, fig.height=2, fig.width=8}
par(mfrow=c(1,4))
invisible(lapply(unique(a[a$drug1=="VU0823511", 'plate.name']), function(pn) do.call(plotGC,append(
    list(ylim=c(-2,5), main=pn, leg=FALSE),getGCargs(a[a$plate.name==pn & a$drug1=="VU0823511",])))))
```

#### Replace drug names
Obtained, from Dave Westover, drug names to replace VU-specific identifiers. 

```{r}
new_drug_names <- readxl::read_xlsx("../data/VLW001_full_drugnames.xlsx")
```

Plan is to replace drug names by matching `d$drug1` to `new_drug_names$VU_ID` and replace with `new_drug_names$"Chemical Name"`. However, the chemical names need to be cleaned.

```{r}
new_drugs <- as.character(new_drug_names$"Chemical Name")
```

Algorithm:

* split on `(`, keeping first part.
* remove dashes
* remove spaces if before numbers
* split on spaces, keeping first part (remove salts)
* If only first letter cap or if != "Ro", make lowercase

```{r}
temp <- sapply(strsplit(new_drugs, " (", fixed=TRUE), "[[", 1, simplify=TRUE)
temp[grep(" [0-9]",temp)] <- gsub(" ", "", temp[grep(" [0-9]",temp)])
temp <- gsub("-", "", temp)
temp <- sapply(strsplit(temp, " "), "[[", 1, simplify=TRUE)
temp[!grepl("Ro", temp) & grepl("^[A-Z][a-z]", temp)] <- 
    tolower(temp[!grepl("Ro", temp) & grepl("^[A-Z][a-z]", temp)])
new_drug_names$drug1 <- temp
```

Replace `d$drug1` values starting with `VU` with appropriate value in `new_drug_names$drug1` then subset data for 2D data only and save into object `a`. 

Exclude `file_name` from Thunor data.
```{r}
d[grep("^VU",d$drug1), 'drug1'] <- 
    new_drug_names[match(d[grep("^VU",d$drug1),'drug1'], new_drug_names$VU_ID),]$drug1

a <- d[grepl("^V2", d$plate.name),]
a <- a[,colnames(a)!="file_name"]
```

#### Save 2D data for Thunor

```{r}
OVERWRITE <- FALSE
thunor_2D_dat <- a
thunor_2D_dat$upid <- paste(thunor_2D_dat$expt.date, thunor_2D_dat$plate.name, sep='_')
thunor_2D_dat$drug1.units <- 'M'
thunor_fn <- "../data/VLW_V23ab_2Donly_Thunor_dataset.csv"
if(!file.exists(thunor_fn) | OVERWRITE) write.csv(thunor_2D_dat, file=thunor_fn, row.names=FALSE)
```

#### Save all data (including file name with full path)

```{r}
alldat_fn <- "../data/VLW_V23ab_dataset.csv"
if(!file.exists(alldat_fn) | OVERWRITE) write.csv(d, file=alldat_fn, row.names=FALSE)
```

